Artech house CDMA mobile radio design jun 2000 ISBN 1580530591 pdf

Viii CDMA Mobile Radio Design2.2 2.2.1 3.1.2 3.1.3 3.2 3.2.1 3.2.2 3.3 3.3.1 The Digital System Architecture IssuesThe MCUThe DSPMemoryMCU FunctionsProtocol AdministrationPower Managemen

CDMA Mobile Radio Design

CDMA Mobile Radio Design

Library of Congress Cataloging-in-Publication Data

Groe, John B.

CDMA mobile radio design/John B Groe, Lawrence E Larson

p cm - (Artech House mobile communications series)

Includes bibliographical references and index

ISBN l-58053-059-1 (alk paper)

1 Code division multiple access 2 Cellular telephone systems 3 Mobile

communication systems I Larson, Lawrence E II Tide III Series

1 Cellular radio Design 2 Wireless communication systems

-Design 3 Code division multiple access

I Tide II Larson, Lawrence E

621.3’845

ISBN l-58053-059-1

Cover design by Igor Valdman

0 2000 ARTECH HOUSE, INC.

685 Canton Street

Norwood, MA 02062

All rights reserved Printed and bound in the United States of America No part of this bookmay be reproduced or utilized in any form or by any means, electronic or mechanical, includingphotocopying, recording, or by any information storage and retrieval system, without permis-sion in writing from the publisher

All terms mentioned in this book that ate known to be trademarks or service marks havebeen appropriately capita&d Artech House cannot attest to the accuracy of this information.Use of a term in this book should not be regarded as affecting the validity of any trademark orservice mark

International Standard Book Number: l-58053-059-1

Library of Congress Catalog Card Number: 00-027455

1 0 9 8 7 6 5 4 3 2 1

Path LossMuitipath FadingModeling the Communication ChannelWireless Standards

References

778141619

2.1 Direct-Sequence Spread-Spectrum

Viii CDMA Mobile Radio Design

2.2 2.2.1

3.1.2

3.1.3

3.2 3.2.1

3.2.2

3.3 3.3.1

The Digital System

Architecture IssuesThe MCUThe DSPMemoryMCU FunctionsProtocol AdministrationPower ManagementDigital Signal Processing AlgorithmsThe Sampling Theorem

Sample Rate ConversionDigital Filters

Fast Fourier TransformsWindowing OperationsDetection Process

References

Speech Coding

Characteristics of Human SpeechSpeech-Coding AlgorithmsWaveform Coders

VocodersSpeech Coders for WirelessCommunication SystemsSpeech Quality

References

292934383940

43

44

4 445

4 646

4 7474949525557586064

6768697072

828385

5.1 5.1.1 5.1'2 5.1'3.

5.2 I

5.2.2 5.2.3 5.2.4 5.2.5

6

61 6.1'1 6.1.2 6.2 6.2'1 6.2'2 6.2'3 6.2'4.

6 3 6.3'1 6.3'2.

6 4 6.4'1 6.4'2.

7

71 7.1'1.

Digital Modem

Digital ModulatorSynchronizationChannel CodingSignal FilteringDigital DemodulatorPilot AcquisitionCarrier RecoverySignal LevelingData DetectionData RecoveryReferences

Data Converters

A/D ConversionIdeal Sampling ProcessNonideal EffectsA/D Converter ArchitecturesParallel A/D ConvertersMultistage A/D ConvertersAlgorithmic A/D ConvertersNoise-Shaping A/D ConvertersD/A Conversion

Ideal ProcessNonideal EffectsD/A Converter ArchitecturesScaling D/A Converter ConceptsOversampled D/A ConvertersReferences

RF System Fundamentals

RF Engineering ConceptsDuplex Operation

8787889194100101103106109113118

121

122122126127128129132134140140141145145146146

149

150150

X CDMA Mobile Radio Design

7.2'2

7.2.3

7 3 7.3'1

7.3'2

7.3'3

7.3.4

7.4 7.4'1

7.4'2

7.4'3

7.4'4

881 8.1'1

SAW Filter TechnologyPower Amplifiers for TransmitterApplications

PA Design Specifications

PA Design TechniquesDevices for PAsReferences

Baseband Channel Select Filters 240

Concepts of Next-Generation CDMA

Next-Generation CDMA and the Physical

Single-Carrier CDMA Option

Forward Link in the Single-Carrier Option

Reverse Link of Single-Carrier Option

Acquisition and Synchronization

Fast Power Control

Air Interface for the Single-Carrier Option

TDD CDMA Option

Multicarrier CDMA Option

Forward Link for the Multicarrier Option

Reverse Link of the Multicarrier Option

252 252 253 257 261 266 266 267 268 270 273 274 276 277 278 279 281

xii CDMA Mobile Radio Design

10.4.3

1111.1

Advanced CDMA Mobile Radios

Advances in Digital Signal ProcessingDSP Performance

Improvements to the Digital ReceiverAdvanced RF Receivers

Image Rejection TechniquesDirect Conversion ReceiversDigital IF Receivers

Comparison of Advanced RF ReceiverArchitectures

Advanced RF Transmitters.Direct Conversion TransmittersSSB Techniques

Predistortion Techniques for AmplifierLinearization

Feedforward PAsLinearized PAs With Nonlinear CircuitsAdvanced Frequency SynthesizersReferences

Glossary About the Authors Index

282283

285

285286287294294298301304304305306

308

3 1 1313317321

325 331 333

Wireless communications is growing at a phenomenal rate From 1991 to

1999, the number of subscribers increased from about 25 million to over 250million Incredibly, over the next seven years, the number of subscribers isexpected to quadruple, to over 1 billion [ 1] That growth rate is faster thanthat of any other consumer electronics product and is similar to that of theInternet

Originally, wireless communications were motivated by and intended formobile voice services Later on, the first analog systems were improved withdigital techniques, providing increased robustness and subscriber capacity Inthe near future, digital systems will be augmented to try to meet users’ insatiableneed for even greater capacity and high-speed mobile data services

Wireless communications rely on multiple-access techniques to sharelimited radio spectrum resources These techniques, which use frequency, time,and power to divide the precious radio spectrum, are described in standardsand are highly regulated As such, infrastructure and subscriber manufacturerscan be different and interchangeable

This book details the complete operation of a mobile phone It describescode division multiple access (CDMA) design issues but presents concepts andprinciples that are applicable to any standard The book emphasizes CDMAbecause next-generation standards are based on that multiple-access technology.This book uniquely ties together all the different concepts that form themobile radio Each of these concepts, in its own right, is suitable material for

a book, if not several books, but is presented in such a way as to highlightkey design issues and to emphasize the connection to other parts of the mobileradio

Xiii

xiv CDMA Mobile Radio Design

Chapter 1 introduces some fundamentals of wireless communications Itdescribes the wireless network, which interfaces with landline services, and theprocedures for communicating through the network Chapter 1 illustrates theeffects of radio propagation and reveals its impact on the mobile phone Italso lists some familiar wireless standards Chapter 2 provides an overview ofCDMA It presents the basic concepts and highlights the key air interfacerequirements for the CDMA IS95 standard

Chapter 3 introduces the digital system, which consists of a digital signalprocessor (DSP) and a microcontroller unit (MCU) The chapter uncovers themyriad of important roles the digital system plays It also reviews some digitalsignal processing fundamentals and describes some tradeoffs in architecture.Chapter 4 introduces speech coding, a key function of the digital system Itshows how voice signals are translated to low bit rate data streams and viceversa Chapter 5 provides detailed information about digital modulation anddemodulaton It presents a practical Rake receiver and describes the receiver’soperation in the network It also points out key timing issues and their effects

on the performance of the mobile phone in the wireless network

Chapter 6 describes data converters, circuits that interface- the digitalsystem to the auditory transducers (microphone and speaker) and the radiofrequency (RF) transceiver The chapter analyzes the nonideal effects of theseinterfaces and also presents fundamental data conversion techniques

Chapter 7 is the first of three chapters dedicated to the RF transceiver,the mobile radio’s connection to the air interface It describes both the RFtransmitter and the receiver from a system perspective, providing critical infor-mation about gain distribution and signal integrity The chapter also presentsinsight into frequency synthesis and frequency planning in the mobile radio.Chapter 8 details the RF transmitter It describes the transmit circuits betweenthe digital-to-analog (D/A) converters’ outputs and the antenna The chaptercovers the I/Q modulator, variable gain amplifier (VGA), up-converter, filters,driver, and power amplifier (PA) Chapter 9 details the operation of the RFreceiver It provides a circuit level view of the receiver from the antenna tothe A/D converters’ inputs This chapter covers the low-noise amplifier (LNA),mixer, VGA, I/Q demodulator, and filters

Chapter 10 describes next-generation wireless services and standards Thechapter points out improvements to CDMA IS95 that will accommodate moreusers and higher data rates It also details leading next-generation CDMAproposals Chapter 11 illustrates architecture advances to support improvedCDMA IS95 pe rformance and to meet the demands of next-generation CDMAnetworks It addresses key areas, including the DSP, the RF transmitter, andthe RF receiver

A book covering such a range of systems, architectures, and circuits crossesseveral engineering disciplines As a result, we benefited from discussions with

Preface xv

and reviews by several colleagues We would like to acknowledge Mr TomKenney, Ryan Heidari, Sassan Ahmadi, and Ken Hsu of Nokia Mobile Phones;Professor George Cunningham of New Mexico Technical University; ProfessorBehzad Razavi of the University of California-Los Angeles; Professors Lau-rence Milstein, Peter Asbeck, Anthony Acompora, and Ian Galton of theUniversity of California-San Diego; Professor John Long of the University

of Toronto; and Mr David Rowe of Sierra Monolithics

Reference

[I] Viterbi, A J., CDMA: Principles of Spread-Spectrum Communications, Reading, MA:

Addison-Wesley, 1795.

Introduction to Wireless

Communications

Wireless technology offers untethered service, newfound freedom, and thepotential for “anytime, anyplace” communications Consumers are embracingthese services enthusiastically; their numbers are growing at a phenomenal rateand will continue to do so, as illustrated in Figure 1.1 The growth andthe excitement of wireless communications are being driven by technologicaladvancements that are making portable radio equipment smaller, cheaper,and more reliable Those advancements include improved signal processing

2 CDMA Mobile Radio Design

techniques, innovative digital and radio frequency (RF) circuit design, andnew large-scale integrated circuit methods

This chapter introduces and describes key aspects of wireless networks

It investigates the wireline backbone, which facilitates wireless communications.That leads to an overview of the communication procedures used by bothwireline and wireless networks The chapter also details the effects of the radiolink, which complicates radio design and leads to a variety of wireless standards

1.1 Network Architecture for Cellular Wireless

Communications

The wireless network supports over-the-air communications between mobileradios and stationary transceivers’ known as base stations These links arereliable only over short distances, typically tens of meters to a few kilometers

As such, a network of base stations is needed to cover a large geographic area,for example, a city Base stations communicate through mobile switchingcenters, which connect to external networks such as the public telephoneswitching network (PTSN), the integrated services digital network (ISDN),and the Internet, as shown in Figure 1.2

The mobile radio is free to move about the network It relies on radiosignals to form a wireless link to the base stations and therefore requires an

RF transceiver To support modern communication methods, the mobile radio

Mobile radio

0 < Mobile radio

Base station

Public telephone switching network, Internet

Figure 1.2 Wireless network architecture is an interconnection of mobile radios, base

stations, mobile switching centers, and the external network.

1 Transmitter-receiver combinations.

Wire& Communications 3

includes a microcontroller unit (MCU) and a digital signal processor (DSP)

to condition the signal before transmission and to demodulate the receivedsignal (Figure 1.3)

The base stations translate the radio signals into data packets and signalingmessages that are readable by the wireline network, which then forwards theinformation to the mobile switching center

The mobile switching center routes the data packets based on the signalingmessages and typically does not originate messages In some cases, the mobileswitching center may need to send queries to find wireless subscribers orportable local numbers (800- and 888-numbers)

The external network provides the communications backbone that nects the mobile switching centers It routes data packets, screens messages forauthorization, verifies routing integrity, and converts protocols The externalnetwork may also act as a gateway to different networks

con-The mobile switching center and the external network are signal transferpoints that include measurement capabilities to indicate network problems and

to monitor usage for billing purposes Built-in redundancies in the networkallow rerouting around faulty network points

The network also includes service control points that interface to ers and provide database access For example, the mobile switching center uses

comput-a service control point to comput-access the home loccomput-ation register (HLR), the visitorlocation register (VLR), and the operation and maintenance center (OMC)files Those databases list the subscribers in the home service area, track any

4 CDMA Mobile Radio Design

roaming (i.e., visiting) subscribers in the coverage area, and hold authenticationfiles

More information on network architectures can be found in [2-4]

1.2 Data Communication Techniques

Modern wireline and wireless networks rely on digital techniques for efficientcommunications The techniques format message signals into data packets,thereby allowing multiple users to be “bundled’ at higher network levels That

is important because it reduces the number of physical connections required

to connect a set of users The bundling occurs at signal transfer points andtypically uses time multiplexing methods [2]

A basic wireline telephone channel for a single user supports a data rate

of 64 Kbps; digital and optical data trunks carry higher data rates, as listed inTable 1.1

The data packets are routed through the network by either circuit-switched

or packet-switched connections In circuit-switched networks, the path betweenthe user and the destination node is set up at the time the connection isestablished, and any needed resources are reserved until the connection isterminated In packet-switched networks, the path is not fixed but is dynamicallyselected based on network loading conditions and the destination addressappended to each data packet

Circuit-switched networks provide dedicated connections with lowlatency, while packet-switched networks offer greater flexibility with improvedefficiency Packet-switched networks are more complicated because data packetscan take different paths and can be received out of order; the data packetsmust then be reassembled prior to final delivery to the user

Table 1.1 Common Data Rates for Digital and Optical Networks [21

Wireh Communications 5

1.3 Protocols for Wireless Communications

Multiple users in communication networks are organized using routing andflow control procedures, known as protocols A protocol is a set of rulesgoverning data transmission and recovery in communication networks Therules ensure reliable, seamless transmission of data and provide network manage-ment functions

Protocols usually are organized as layers in a communication “stack.”Data is passed up or down the stack one layer at a time, with specific functionsperformed at each layer

Most communication networks follow the open system interconnections(OSI) model [5] The seven-layer protocol stack, shown in Figure 1.4(a),includes the physical, data link, network, transport, session, presentation, andapplication layers In wireless communication networks, a variation of the OS1model, the signaIing system number 7 (SS7) model [2-31, is used This four-level protocol stack, shown in Figure 1.4(b), mirrors the first three layers ofthe OS1 model and combines the higher levels into a single application layer.The protocol stack defines the architecture of each signal transfer point

or node in the network It uses the physical layer to interconnect those nodesand provide a path through the network, plus the data link and network layers

to translate control signals and reformat data for communication with different

Figure 1.4 Network models: (a) OSI protocol stack typical of wireline networks and (b)

SS7 protocol stack followed by wireless networks.

6 CDMA Mobile Radio Design

networks Data always flows from one layer to the next in the protocol stack,

as shown in Figure 1.5, to ensure robust communications

Each layer in the protocol stack performs essential operations that aredefined by the topology of the communication network Those operations areoutlined next

The physical layer is the interface between two communication nodes

In a wireless network, the physical layer is the air interface between the mobileterminal and the base station In a typical wireline network, it is the digital

or optical trunk The physical layer provides transfer services to higher layers

in the protocol stack Those transfer services use physical channels, also known

as transport channels, with defined data rates, modulation schemes, powercontrol methods, and RF parameters The physical layer is different for eachunique communication standard

The data link layer combines the medium access control (MAC) andradio link control sublayers The MAC sublayer maps basic functions known

as logical channels to physical channels That can be straightforward, or it caninclude multiplexing several logical channels onto a common physical channel.The data link layer also provides message sequencing, traffic monitoring, andsignal routing to higher protocol layers

The radio link control sublayer breaks down the data stream into datapackets, also known as transport blocks, for transmission It includes errorcontrol to ensure the integrity of the transmitted data Typically, that means

a parity check or a cyclic redundancy check (CRC) based on a polynomialgenerator [6] The radio link control layer also interfaces with the higherprotocol layers and provides call initialization, connection, and termination.The network layer (or radio resource control layer) provides control andnotification services It supervises radio resources, including physical channelassignments, paging requests, and transmit power levels It also interfaces tothe wireline network and thereby enables connections to other users

Network path Network

Wirehs Com~unicatiom 7

The application layer represents the destination node It specifies of-service (QoS) requirements (priority levels, security, response time expecta-tions, error rates, and recovery strategies) without the restrictions of the airand network interfaces The application layer compresses and expands data intime to match the expectations of the mobile user

quality-The physical layer, the data link layer, and the network layer combine

to form the message transfer part (MTP) of the SS7 protocol stack, as shown

in Figure 1.6 The MTP of the SS7 model covers transmission from node tonode in the communication network It also interfaces with high-level protocolstailored to specific applications For voice communications, one of two high-level protocols is used: the telephone user part (TUP) or the ISDN user part(ISUP)

1.4 Radio Propagation in a Mobile Wireless Environment

The radio interface is unique to wireless communications and is responsiblefor much of the complexity associated with wireless networks and mobilephones The radio interface between the mobile phone and the base station

is referred to as the communication channel and is affected by large- andsmall-scale factors The large-scale effects are due to simple attenuation of thetransmitted signal through the atmosphere The small-scale effects behaveunpredictably, vary sharply over small distances, and change quickly

Figure 1.6 The SS7 model and the relationships among its constituent parts.

8 COMA Mobile Radio Oesign

where r(d) is the received power at a distance d separating the mobile and the

base station, and n is the path loss exponent with typical values of 2.7 to 3.5for urban cellular radio [7] The model is quite simple and is appropriate onlyfor line-of-sight propagation

In practice, the signal path typically is cluttered by obstructions thatreflect or block the transmitted signal and introduce statistical variability tothe simple path loss model, as shown in Figure 1.7 This effect is known asshadowing and is modeled as a log-normal random variable [7] That leads to

a new expression for the received power:

r(d) 0~ lox/lo&-n (1.2)where x is the log-normal random variable used to model the shadowing effect

d

Figure 1.7 Received signal strength with path loss and log-normal shadowing.

Introduct& to Wi’rcks Communications

Figure 1.8 Multipath propagation of a transmitted signal arrives at the receiver with

different delays.

Elapsed Time (mS)

Figure 1.9 Multipath fading produces a wide variation in the received signal strength as

a function of time in a mobile environment (from: T S Rappaport, Wireless Communicarions, 0 1995; reprinted by permission of Prentice-Hall, Inc., Upper Saddle River, NJ.)

10 CDMA Mobile Radio Des@

destructively, and the received signal can disappear completely for a shortperiod of time

The effects of multipath fading combine with large-scale path losses toattenuate the transmitted signal as it passes through the channel, as shown inFigure 1.10 The graph shows that the received power level at a distance d

from the transmitting antenna depends on the simple path loss model altered

by the shadowing and multipath distributions

Multipath fading is created by the frequency-selective and time-varyingcharacteristics of the communication channel Those characteristics are notdeterministic and therefore must be analyzed using statistical methods Thisapproach is illustrated in the following examples

In the first case, two sinusoidal signals at frequencies fl and f2 aretransmitted through the channel as shown in Figure 1 l 1 The signals areaffected by the channel, which attenuates the power level, T, of each signalindependently The attenuation process for each signal varies with frequencyand can be described by two distinct probability density functions (pdf’s) If

fi = f2, then the pdf’s o tf h e received power levels p (7) will be nearly thesame, and the cross-correlation ‘between the two, R(Af ), will be high As theseparation between fi and f2 increases, their amplitude pdf’s will becomedissimilar and their cross-correlation will be lower

The coherence bandwidth, (Af ),,is the range of frequencies in which theresponse of the channel remains roughly constant, that is, the cross-correlation isgreater than one-half In other words, the channel affects a range of frequencies

(Af )C, from fi to fi, similarly

Therefore, narrowband signals that fit within the coherence bandwidth,experience nearly constant, or “flat,” frequency fading That implies that the

d

Figure 1.10 Shadowing and multipath propagation affect received signal strength.

In the second example, two identical signals are transmitted at differenttimes, tl and t2, as shown in Figure 1.12 The channel affects each signal’sreceived power level independently and produces distinct pdf’s for the twooutput waveforms The pdf’s are cross-correlated to reveal changes in thechannel If tl = t2, the cross-correlation of the two waveforms will be high.But as the separation between tl and t2 increases, the cross-correlation willbecome lower and eventually fall below one-half That indicates the timeseparation benveen signals where the channel response stays constant, that is,the time coherence of the channel, (At), In other words, the response of thechannel and the received power level is predictable as long as the separation

in time between signals is less than the time coherence of the channel.The coherence bandwidth and time coherence parameters are key mea-sures of the communication channel These parameters lead to a second set

of parameters, known as the scattering functions, that describe the effect on

2 Most cellular CDMA systems, such as CDMA IS95 and WCDMA, use direct-sequence spread-spectrum modulation.

12 COMA Mobile Radio Design

Path loss shadowing, multipath fading -

Figure 1.12 Time-varying behavior of the channel affects two pulses transmitted at

separate times differently.

the transmitted signal The scattering functions S(T, Y) are found by takingthe Fourier transforms of the cross-correlation functions, that is,

where the multipath delay spread, r, is related to ll(Af)c and the dopplerspread, Y, is associated with l/(At),

The cross-correlation parameters and scattering functions are small-scaleeffects caused by multipath propagation through the communication channel.These multipath rays are duplicate signals that are scaled and phase rotatedrelative to each other Interestingly, at any instant t,, the received signal is acomposite of these replica signals Consequently, the received signal at time

t, is described by

n=Owhere a, is the complex amplitude of the nth multipath rays.

The multipath delay spread (7) is especially important in digital cation systems It measures the smearing or spreading in the received signalwhen an impulse is transmitted through the communication channel Impulsesmearing is shown in Figure 1.13 for a typical cellular system The first peak

communi-in the response generally corresponds to the lcommuni-ine-of-sight ray, while the otherpeaks reveal the scaling and propagation delay of the strong multipath rays.The delay spread covers the time interval from the first peak to the last significantpeak

:4- RMS delay spreadw

x

$>

a>

.-ii -10

?3

N

(d -20

.-E 6 z

-30

Excess delay (mS)

Figtire 1.13 Measured multipath delay spread for a typical cellular system (From: T S.

Rappaport, Wireless CommunicaG~ns, 0 1995; reprinted by permission of Prentice-Hall, Inc., Upper Saddle River, NJ.)

The delay spread causes adjacent data bits to overlap and producesintersymbol interference (ISI) I n narrowband communication systems, thatcan be disastrous and must be removed by equalization techniques In widebandsystems, it is possible to remove the delay from the multipath componentsand to align the rays using signal processing methods.3 That yields the ensembleaverage of the received power,

3 The most common approach to aligning the rays and constructively summing them is the Rake receiver, which is described in Chapter 5.

14 CDMA Mobile Radio Design I

1.4.3 Modeling the Communication Channel

The wireless communication channel is unpredictable, making deterministicmodels of performance impossible [7-lo] As a result, the performance ofwireless communication systems is assessed using simplifications of practical

or particularly troublesome environments based on three basic models

Figure 1.14(a) illustrates the simplest propagation model, line-of-sightpropagation in a noisy environment Here, the received signal is given by

where c is the path loss factor, s(t) is the transmitted signal, and n (t) is the

added channel noise The noise is constant over frequency and is usually referred

to as white noise, while its amplitude is described by a zero-mean Gaussianpdf The function is defined by

(1.7)

2 '

where CT 1s the variance of the random variable a This type of noise source

is called additive white Gaussian noise (AWGN) The line-of-sight model isappropriate for picocells or for wireline communications

Wireless communication channels, however, are both time varying andfrequency dependent Therefore, the path loss factor of the line-of-sight model

is altered to provide for the variation with time t and excess delay T 4 T h ereceived signal is then

r t = c(t, 7) l ( ) + n t

( W

where c(t, 7) is a function that describes the wireless channel and models bothlarge-scale and small-scale effects By contrast, the line-of-sight model assumesthat c is constant

This second, improved model of the wireless channel is approximated inthe following way A signal cosot is transmitted via the wireless channel and

received at the receiver as rcos (w t + +), where r is a complex amplitude and

4 is a uniformly distributed random variable The complex amplitude r can

be modeled as independent I and Q random variables [8] Furthermore, thereare a sufficient number of independent reflections (multipath rays) to allowthose random variables to be modeled as Gaussian distributed with

4 The excess delay spread is tied to the coherence bandwidth R(Af )c.

Wirehs Communications

A W G N

PM

A W G N

Figure 1.14 Channel models: (a) line of sight with AWGN, (b) Rayleigh channel model,

and (~1 Rician channel model.

The probability of receiving a signal of amplitude r follows a Rayleigh

or Rician distribution that depends on the mean of the random variables Iand Q If the mean of both random variables is zero, the pdf of T is Rayleighdistributed and equal to

16 COMA Mobile Radio Design

(1.9)

where a2 is the time-averaged power level That produces the channel model

shown in Figure 1.14(b) If the mean of the random variables is nonzero, a

dominant multipath component or a line-of-sight path is present and the pdf

is Rician, that is,

(1.10)

where A is the peak of the dominant signal and lo(*) is the modified Besselfunction of the first kind and zero order That leads to the channel modelshown in Figure 1.14(c)

The Rician factor k describes the strength of the line-of-sight ray andequals

k =- A2

As k approaches infinity, the Rician distribution becomes a delta function,which matches the simple line-of-sight model As k approaches zero, the Riciandistribution transforms into a Rayleigh distribution

The AWGN, Rayleigh, and Rician channel models are simple, compactmodels for approximating the effects of radio propagation An overview ofmore complicated models is available in [ 1 I]

To limit interference, frequency channels are generally assigned based on the

A(f) Introduction to Wireh Communications

f, f* i f3 ; f

* i&- Channel (a)

Figure 1.15 Multiple access methods: (a) frequency division multiple access (FDMA),

(b) time division multiple access (TDMA), and (c) code division multiple access (CDMA).

18 CDMA Mobile Radio Design

frequency reuse pattern shown in Figure 1.16 In special cases, such as CDMA

networks, universal frequency reuse is allowed and is a powerful advantage.The choice of multiple-access technique directly affects subscriber capac-ity, which is a measure of the number of users that can be supported in apredefined bandwidth at any given time

First-generation (1 G) wireless communication systems use analog

meth-ods These systems superimpose the message signal onto the RF carrier usingfrequency modulation (FM) and separate users by FDMA techniques Anexample of this type of system is the Advanced Mobile Phone System (AMPS).Second-generation (2G) communication systems introduce digital tech-nology These systems digitally encode the message signal before superimposing

it onto the RF carrier Digital data allows powerful coding techniques thatboth improve voice quality and increase network capacity Examples of thistype of system include GSM (Global System for Mobile Communications) [ 121,NADC (North American Digital Cellular) [ 131, PHS (Personal HandyphoneSystem) [ 141, and CDMA IS9 5 [ 153

Table 1.2 compares some of the leading wireless standards

+ Cell separation +I Figure 1.16 Seven-cell reuse pattern typically used by carriers to separate frequency

channels.

Table 12 Important Properties of Some Leading Wireless Standards

FDMA 30K FM NA 1 47

880-915 824-849 925-960 869-894 1,710-1,785 1,850-1,910 1,805-l ,885 1,936-l ,990 FfiDMA FJTDMA

v/4DQPSK

8 0 m W

32 Kbps 4 19

824-849 869-894 1,850-1,910 1,930-l ,990 F/CD MA 1.25M QPSK 200mW l-8 Kbps 28 224

Dataquest Survey of Worldwide Wireless Subscribers, Nov 1999.

Modarressi, A R., and R A Skoog, “Signaling System No 7: A TutoriaI,” IEEE

Communications Magazine, July 1990, pp 19-35.

Russel, T., Signaling System #7, New York: McGraw-Hill, 1998.

Gallagher, M D., and R A Snyder, Mobik Tekcommunications Networking, New York:

Steele, R., ed., Mobile Radio Communications, New York: IEEE Press, 1992.

Proakis, J G., Digital Communications, New York: McGraw-Hill, 1995.

Anderson, J B., and T S Rappaport, “Propagation Measurements and Models for

20 CDMA Mobile Radio Design

[13] TWEIA Interim Standard, “Cellular System Dual Mode Mobile Station-Base StationCompatibility Standard,” IS-54B, Apr 1992

[14] Personal Handiphone System RCR Standard 28, Ver 1, Dec 20, 1993

[15] TWEIA Interim Standard, “Mobile Station-Base Station Compatibility Standard forDual-Mode Wideband Spread Spectrum Cellular System,” IS-9SA, Apr 1996

The CDMA Concept

CDMA is a multiple-access scheme based on spread-spectrum communicationtechniques [l 3] It spreads the message signal to a relatively wide bandwidth

by using a unique code that reduces interference, enhances system processing,and differentiates users CDMA does not require frequency or time-divisionfor multiple access; thus, it improves the capacity of the communication system.This chapter introduces spread-spectrum modulation and CDMA con-cepts It presents several design considerations tied to those concepts, includingthe structure of the spreading signal, the method for timing synchronization,and the requirements for power control This chapter also points out CDMAIS95 [4] details to illustrate practical solutions to these design issues

2.1 Direct-Sequence Spread-Spectrum Communications

Spread-spectrum communications is a secondary modulation technique In atypical spread-spectrum communication system, the message signal is firstmodulated by traditional amplitude, frequency, or phase techniques A pseudo-random noise (PN) signal is then applied to spread the modulated waveformover a relatively wide bandwidth The PN signal can amplitude modulate themessage waveform to generate direct-sequence spreading, or it can shift thecarrier frequency of the message signal to produce frequency-hopped spreading,

2 2 CDMA Mobile Radio Design

The CD&L4 Conqt 23

In most cases, the PN signal is a very high rate, nonreturn-to-zero (NRZ)pseudorandom sequence that chops the modulated message waveform intochips, as shown in Figure 2.2 Hence, the rate of the secondary modulatingwaveform is called the chip rate, fc, while the rate of the message signal isdesignated the bit rate, f6 The two modulation processes produce differentbandwidths, namely, R for the modulated message signal and W for the rela-tively wide spread-spectrum waveform Note that the secondary modulationdoes not increase the overall power of the message signal but merely spreads

it over a wider bandwidth

The frequency-hopped spread-spectrum signal is formed by multiplyingthe message signal with a pseudorandom carrier frequency opn(t):

The direct-sequence spread-spectrum signal formed in a simple and idealtransmitter can be described by

s(t) = pn(t)Ad(t)cos(wt + e) (2.3)where pn(t) is the pseudorandom modulating waveform, A is the amplitude

of the message waveform, d(t) is the message signal with bipolar values +l,

24 CDMA Mobile Radio Design

w is the carrier frequency, and 8 is a random phase The signal is transmittedover the air interface and is received along with thermal noise n(t) and interfer-ence i(t), which are added by the channel The received signal is’

interfer-R Any received interference, i(t), is spread by the correlator to the relativelywide bandwidth W, and its effect is lowered

The correlator affects the message signal d(t) differently than it does theinterference i(t) and thereby improves the signal-to-noise ratio (SNR) of thereceived signal.2 That powerful benefit is the processing gain of the system and

is equal to the spreading factor W/R.

In general, the spreading signal is a binary waveform with values specified

at the chip rate The binary waveform allows easy implementation withoutsacrificing performance and enables synchronization of the transmitter to thereceived signal It is possible to achieve a continuous-time waveform by passingthe binary signal through a linear filter

1 To illustrate the spread-spectrum concept, delay and scaling effects introduced by the channel are ignored here.

2 Noise refers to any unwanted energy and includes interference.

c?lMA Concept 25

These characteristics are available from deterministic, pseudorandomsequences with the following classical properties:

l There are near-equal occurrences of + 1 and -1 chips

l Run lengths of r chips with the same sign occur approximately 2-rtimes

l Shifting by a nonzero number of chips produces a new sequence thathas an equal number of agreements and disagreements with the originalsequence [ 11

The randomness of the signal p(t) is measured by the autocorrelationfunction R,,(T), given by

Ti2 Rp,, (r) = Lim ‘-

Figure 2.3 Autocorrelation of PN sequence.

26 CDMA Mobile Radio Design

The uniqueness of the signal pn (t) is analyzed with the cross-correlationfunction, defined by

1

R?(7) =

Lim-T-+J I x(t)y(t + 7)dt (2.8)-T12

pseudo-The Hadamard code is a commonly used orthogonal code [G] It is based

on the rows of a square (n by n) matrix known as the Hadamard matrix Inthe matrix, the first row consists of all OS, while the remaining rows containequal occurrences of OS and 1s Furthermore, each code differs from everyother code in n/2 places

The Hadamard matrix if formed by the following recursive procedure:

(2.19

Figure 2.4 Cross-correlation of a PN sequence.

The CDMA Concept 27 -

where Wn is derived from Wn by replacing all entries with their complements.The Hadamard matrix provides n orthogonal codes

The noise (Nt) seen by the correlator is the signal energy received fromthe k - 1 users and thermal noise, that is,

28 CDMA Mobile Radio Design

The SNR is a key consideration in all communication systems In digital

communication systems, the SNR is characterized by a related figure of merit,

the bit energy per noise density ratio (Eb /NO) That parameter takes into

account the processing gain of the communication system, a vital consideration

in spread-spectrum communications The parameter normalizes the desired q:signal power to the bit rate R to determine the bit energy and the noise or

interference signal power to the spreading bandwidth Wto determine the noise

spectral density Recall that the correlator

‘-l Despreads or integrates the desired signal to the narrow bandwidth

of the original message signal (R);

l Spreads the interference to a wider bandwidth;

l Leaves the uncorrelated noise unaltered

Therefore,

Eb SIR S m=-=:

Amazingly, the interference from other users (i.e., self-interface) is reduced

by the processing gain (W/R) of the system

A simple expression for the capacity of a CDMA system is developed

from (2.15) and is given by

W/R

where (Eb IN,),i, is the minimum value needed to achieve an acceptable level

of receiver performance, typically measured as the bit error rate (BER) The

expression shows that the capacity of CDMA communication systems depends

heavily on the spreading factor and the receiver’s performance The capacity

is tied to a flexible resource-power-and is said to be sofi-limited In other

words, if the required Eb/NO is lowered, the transmit signal power allocated

to each user is reduced, and the number of users can be increased In contrast,

the capacity of systems that employ other multiple-access methods like FDMA

and TDMA are hard-limited That is because their capacity is fured by system

design